Naked Science Forum

Non Life Sciences => Physics, Astronomy & Cosmology => Topic started by: constant escape on 21/07/2020 01:30:51

Title: Is there a metric for complexity?
Post by: constant escape on 21/07/2020 01:30:51
Hello all. Still new here, and did a brief search but couldn't find a thread oriented around this. If there is, I'd be happy to read through it.

Is there some kind of metric for complexity? For orderedness? Any kind of constant to work with? Not sure even how to define it in physical terms - perhaps something that compounds moles with some measure of heterogeneity? I find myself severely lacking in vocabulary here.

Full disclosure, I come from philosophy, in terms of approach. One of my interests is systems, specifically how systems can combine into systems-of-systems, and how the ontologies correlate between the two scales - if there is a correlation to discern. I'm in way over my head, regarding the science, but hopefully a few deep inquiries (along with plenty of cherrypicking) will suffice. It seems the span of disciplines is too vast to avoid such cherrypicking, but maybe there is a way to do due diligence.

More broadly, I'm interested in how intelligence emerges, and how logos emerges from intelligence. Is there a name for the kind of organization that is antecedent to intelligence? Dynamics of emergence and supervenience, generalized as much as possible, seem to be at work at a number of scales, and I am dying to figure it out.
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 21/07/2020 02:52:32
There are measures for complexity but they are generally heavy on the math side. I think what you might want is to get a feel for what complexity is at the conceptual level.

You might want to start with Algorithmic Information Theory (https://en.wikipedia.org/wiki/Algorithmic_information_theory)  as a source of ideas. It can get pretty hairy technically speaking if you follow the links, which is why I stress taking away ideas rather than specifics.

This paper (https://royalsocietypublishing.org/doi/10.1098/rsos.172221) might also give you some ideas.
Title: Re: Is there a metric for complexity?
Post by: constant escape on 21/07/2020 03:15:57
Thank you. And yes, the shallow research I've done into computational complexity and whatnot has been frighteningly heavy, but, like you said, there can still be ideas to walk away with.

I haven't heard of algorithmic information theory, but it sounds right up my alley.

Any other tips for interdisciplinary research in general?
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 21/07/2020 03:18:49
Any other tips for interdisciplinary research in general?

For now, keep away from anything that uses the word 'entropy'.
Title: Re: Is there a metric for complexity?
Post by: constant escape on 21/07/2020 04:41:27
You mean, it's a buzz word? Or it's too complex to begin with?
Title: Re: Is there a metric for complexity?
Post by: Bored chemist on 21/07/2020 08:51:54
it's too complex to begin with?
Yes.
Title: Re: Is there a metric for complexity?
Post by: evan_au on 21/07/2020 12:28:50
Quote from: OP
I'm interested in how intelligence emerges
That is certainly a complex topic, and many researchers are trying to investigate it by various means (including computational models and by analyzing living and dead brains).
https://en.wikipedia.org/wiki/Mind

Solving it may take a while...
Quote from: Emerson M. Pugh
If the human brain were so simple that we could understand it, we would be so simple that we couldn’t.
https://quoteinvestigator.com/2016/03/05/brain/


Quote
how logos emerges from intelligence
The method of making logos is a mystery, but there are plenty of advertising agencies who will do it for you (for a fee)... ;)

Logos is Greek for word or speech. But what do you mean by it?
Title: Re: Is there a metric for complexity?
Post by: constant escape on 21/07/2020 21:50:40
I mean it, I believe, in as broad a sense as it could be meant. All discourse, anything drawn, written or said. Tough, for me, to boil it down to a single articulation. Rationality would be a major, if not the major, tentpole. Perhaps it can be considered as the expression of intelligence? The voicing of order? Does it necessarily have to do with communication?

Particularly, how logos parcels itself out into domains and subdomains, each taking a unique topic(s) and a unique mode of understanding. What kind of function can we map onto this parcelling? And where would such a function aim? Do these domains close in on perfection, or do they endlessly proliferate into subdomains?
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 22/07/2020 02:27:19
I mean it, I believe, in as broad a sense as it could be meant. All discourse, anything drawn, written or said. Tough, for me, to boil it down to a single articulation. Rationality would be a major, if not the major, tentpole. Perhaps it can be considered as the expression of intelligence? The voicing of order? Does it necessarily have to do with communication?

Particularly, how logos parcels itself out into domains and subdomains, each taking a unique topic(s) and a unique mode of understanding. What kind of function can we map onto this parcelling? And where would such a function aim? Do these domains close in on perfection, or do they endlessly proliferate into subdomains?

Not that I want to shift the conversation in that direction, but are you using logos in the manner of Heraclitus to mean the principle of order and unity? Or the Stoics as the active principle of reason underlying the universe?

I am not looking to discuss philosophy. I just want to get a better feel for what you are looking for in terms of useful material.
Title: Re: Is there a metric for complexity?
Post by: Edwina Lee on 22/07/2020 05:42:34
Complexity is a very general word used in numerous situations. You need to specify the objectiveness of wanting to measure complexity.

For example:-
An empolyer may want to renumerate employees in relation to the complexity of the job;
An assembly project may be rewarded according to the complexity of the assembly work;
A project may need to be priced according to its complexity;
A software development methodology may be evaluated based on how complexity may increase;

. . . .
So you see, you can't quantify complexity without specifying what the purpose of the quantification is.
Title: Re: Is there a metric for complexity?
Post by: evan_au on 22/07/2020 11:00:15
Quote from: constant escape
do they endlessly proliferate into subdomains?
One of the more complex areas to analyse is the area of turbulence.
- Turbulence has eddies
- Which spawn smaller eddies
- Which spawn even smaller eddies
- Which all becomes horribly complex

But in essence, the behaviour on very tiny scales affects the overall behavior on large scales.
- This is one of the unsolved problems in physics
- And there is a $1 million prize hanging on a solution to one part of the problem (around the Navier-Stokes equations)

See: https://en.wikipedia.org/wiki/Turbulence#Examples_of_turbulence
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 22/07/2020 17:03:28
Quote from: constant escape
do they endlessly proliferate into subdomains?
One of the more complex areas to analyse is the area of turbulence.
- Turbulence has eddies
- Which spawn smaller eddies
- Which spawn even smaller eddies
- Which all becomes horribly complex

But in essence, the behaviour on very tiny scales affects the overall behavior on large scales.
- This is one of the unsolved problems in physics
- And there is a $1 million prize hanging on a solution to one part of the problem (around the Navier-Stokes equations)

See: https://en.wikipedia.org/wiki/Turbulence#Examples_of_turbulence

Bigger whorls have little whorls
That feed on their velocity
Little whorls have lesser whorls
And so on to viscosity

The Millennium Problem I would love to see solved is Yang-Mills and the  Mass Gap (https://www.claymath.org/millennium-problems/yang%E2%80%93mills-and-mass-gap).  It would tell us more about how the universe works and make it less messy and more orderly.
Title: Re: Is there a metric for complexity?
Post by: constant escape on 22/07/2020 17:23:04

Not that I want to shift the conversation in that direction, but are you using logos in the manner of Heraclitus to mean the principle of order and unity? Or the Stoics as the active principle of reason underlying the universe?

I am not looking to discuss philosophy. I just want to get a better feel for what you are looking for in terms of useful material.

I would do well to look into both of those, because I can't say I'm familiar enough to elaborate upon them.

In trying to overlap the various translations of logos, it seems that it is manifest in words, but it is not limited to such manifestation. As you said, we need not totally veer into this territory, but for the sake of clarification, I think that's what I mean: Logos as a methodology/order that manifests as, most essentially, words.

I can certainly dig deeper into literary theory, but that doesn't seem to get to the core, from what I gather. Maybe philology? Again, not sure.

Maybe it would require a more general familiarity with emergence, period. Could logos be a function (?) of a high enough quantity of neurons?

To be honest, and perhaps logos just isn't the right word for it - I would consider animal languages to be logos. Whatever I'm talking about, whether or not it is logos proper, is not limited to humans, but is most robustly manifest through humans.
Title: Re: Is there a metric for complexity?
Post by: Bill S on 23/07/2020 21:37:13
Quote
Bigger whorls have little whorls
That feed on their velocity
Little whorls have lesser whorls
And so on to viscosity

Like it!     

But once it cools and sets so thick
It waxes really sticky,
Internal friction plays its part.
Escape can be quite tricky.
Title: Re: Is there a metric for complexity?
Post by: alancalverd on 24/07/2020 13:02:57
The Millennium Problem I would love to see solved is Yang-Mills and the  Mass Gap.  It would tell us more about how the universe works and make it less messy and more orderly.
Possibly off topic but maybe worth considering:

The inherent intellectual discomfort of the Big Bang is the idea of "something from nothing", that all the mass and energy of the universe appeared spontaneously with no precursor. Cyclic universes have some appeal but still seem to involve uncomfortable singularities that don't fit with observed physical laws.

But suppose there are truly complementary entities with negative mass. We then have a "true universe" T with zero mass/energy, consisting of the observable universe U and an equivalent complement V.

U particles ua, ub.... interact as observed, but the interaction of a u and a v  particle is very different. They cannot coalesce because ultimately the gravitational force Gmumv/r2 is negative and unlimited, so unlike "conventional" uelectron and upositron antimatter interactions, for instance, they cannot mutually annihilate. But they can force each other apart, and the presence of any v particle between two u particles will result in the observable dispersal of the u's.

Hence the inevitable observed expansion of U.

We cannot observe any v, even though we can hypothesise its behavior. A v electron will have the same charge as a u electron, so will move in the same direction in an electric field, but as it will not interact with any u particle, we cannot detect it with U instruments!
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 24/07/2020 16:56:43
The Millennium Problem I would love to see solved is Yang-Mills and the  Mass Gap.  It would tell us more about how the universe works and make it less messy and more orderly.
Possibly off topic but maybe worth considering:

The inherent intellectual discomfort of the Big Bang is the idea of "something from nothing", that all the mass and energy of the universe appeared spontaneously with no precursor. Cyclic universes have some appeal but still seem to involve uncomfortable singularities that don't fit with observed physical laws.

But suppose there are truly complementary entities with negative mass. We then have a "true universe" T with zero mass/energy, consisting of the observable universe U and an equivalent complement V.

U particles ua, ub.... interact as observed, but the interaction of a u and a v  particle is very different. They cannot coalesce because ultimately the gravitational force Gmumv/r2 is negative and unlimited, so unlike "conventional" uelectron and upositron antimatter interactions, for instance, they cannot mutually annihilate. But they can force each other apart, and the presence of any v particle between two u particles will result in the observable dispersal of the u's.

Hence the inevitable observed expansion of U.

We cannot observe any v, even though we can hypothesise its behavior. A v electron will have the same charge as a u electron, so will move in the same direction in an electric field, but as it will not interact with any u particle, we cannot detect it with U instruments!

I keep trying to put together a full explanation of my take on a double sided universe, connected at the origin. One side would be positive mass-energy, the other negative mass-energy.  The positive mass-energy would shape spacetime in one time direction, The negative mass-energy would shape spacetime in the other time direction. Both sides would have the same laws, thanks to CPT symmetry, and don't forget that the T stands for time reversal symmetry. But that all belongs in another thread.

BTW positive mass-energy matter and negative mass-energy antimatter of the corresponding type would interact. They would annihilate to nothing. Everything would cancel.
Title: Re: Is there a metric for complexity?
Post by: alancalverd on 24/07/2020 23:38:36
positive mass-energy matter and negative mass-energy antimatter of the corresponding type would interact.
Only if they could approach and coalesce, which is prevented by their mutually repulsive gravitation.
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 25/07/2020 02:30:25
positive mass-energy matter and negative mass-energy antimatter of the corresponding type would interact.
Only if they could approach and coalesce, which is prevented by their mutually repulsive gravitation.

If you threw a negative mass-energy anti-proton at a positive mass-energy proton, gravity would not be significant factor. Normal protons repel each other despite being attracted gravitationally.

But we are getting way off topic.
Title: Re: Is there a metric for complexity?
Post by: alancalverd on 25/07/2020 12:47:13
The existence of black holes is evidence that "positive" gravitation ultimately dominates electrostatic repulsion, so it is reasonable to assume that negative gravitation will equally prevent the coalescence of electrically attractive particles of opposite mass.   
Title: Re: Is there a metric for complexity?
Post by: Malamute Lover on 26/07/2020 00:24:01
The existence of black holes is evidence that "positive" gravitation ultimately dominates electrostatic repulsion, so it is reasonable to assume that negative gravitation will equally prevent the coalescence of electrically attractive particles of opposite mass.

We are way outside the thread topic now. I will put together my magnum opus :D :D :D when I can. RL and earlier threads keep getting in the way.